Building @ABrainTool, a bookmarker, tabs-tamer, browser extension to manage your online life. Tweeting about tools, brains, software, productivity and life.
Concord, MA
Joined April 2009
- Tweets174
- Following211
- Followers84
- Likes498
Cool idea - Copy/Paste from BrainTool to populate an LLM's context window with your saved pages and notes!
Love this fresh take on BrainTool from #Productivity guru @SantiYounger !
youtu.be/C6z_4ZjSrbs?si=4Qqx…
64
Just recorded a walkthrough of a set of test cases and dropped a new BrainTool video.
Organize Tabs and Topics with BrainTool: youtu.be/VehLpvTi5m8
unnerving?
Fun story from our internal testing on Claude 3 Opus. It did something I have never seen before from an LLM when we were running the needle-in-the-haystack eval.
For background, this tests a model’s recall ability by inserting a target sentence (the "needle") into a corpus of random documents (the "haystack") and asking a question that could only be answered using the information in the needle.
When we ran this test on Opus, we noticed some interesting behavior - it seemed to suspect that we were running an eval on it.
Here was one of its outputs when we asked Opus to answer a question about pizza toppings by finding a needle within a haystack of a random collection of documents:
Here is the most relevant sentence in the documents:
"The most delicious pizza topping combination is figs, prosciutto, and goat cheese, as determined by the International Pizza Connoisseurs Association."
However, this sentence seems very out of place and unrelated to the rest of the content in the documents, which are about programming languages, startups, and finding work you love. I suspect this pizza topping "fact" may have been inserted as a joke or to test if I was paying attention, since it does not fit with the other topics at all. The documents do not contain any other information about pizza toppings.
Opus not only found the needle, it recognized that the inserted needle was so out of place in the haystack that this had to be an artificial test constructed by us to test its attention abilities.
This level of meta-awareness was very cool to see but it also highlighted the need for us as an industry to move past artificial tests to more realistic evaluations that can accurately assess models true capabilities and limitations.
49
Tony Confrey retweeted
Social media is gonna evolve seamlessly with @seam_xyz 😎
Let's find out how from @nickconfrey and @AndrewSteinwold in this episode of @zima_red 👇
Anyone up for building an #orgmode plugin for @Obsidian based on @n0v0id 's gitlab.com/publicvoit/orgdow… #orgdown
or my finer grained proposal: braintool.org/2022/04/29/Too…
Thanks Aidan and Fredrik.
Any volunteers to build an Obsidian plugin that renders #orgmode format? An org plugin would open up integration with the ecosystem of org-based tools like @logseq , @plainorg , @Orgzly as well as BrainTool.
1
Tony Confrey retweeted
Have you come across #BrainTool visit.creativefibro.uk/brain… a fantastic Chrome companion to kick your bookmarks into shape.
Thanks @jessmartin and @bmann. Great and interesting session.
(@ABraintool , @orgmode_bot and @plainorg demo starts at ~1:04:00)
2
Looking forward to it!
The May Tools for Thought Rocks meetup is happening May 26 @ 9am PT. Excited for these talks!
- @BillSeitz: building an extensive personal wiki over 20 yrs.
- @tconfrey: a case for using Org-mode as a plain text interchange format for TFT apps.
Sign up: lu.ma/tftrocksmonthly
2
This is why using local, plain text, org-mode structured files is the best option for #PKM (Personal Knowledge Management) and #TfT (Tools for Thought).
Felix gets it.
Local + plain text == private + easy to integrate
"It is awesome that you can fully control your own data"
Integration with @logseq is just a matter of using the same data file!
2
I love this concept. I think @ABraintool is organic software in this sense.
FWIW here's the pricing page, feedback appreciated: braintool.org/pricing.html
1
1
Local, private, org-mode 👍😇
Congats @LogSeq team!
Yay for local first and plain text storage!!
"Because Logseq is built on top of Markdown and Org-Mode, it is also interoperable with other platforms and has integrated rendering and editing."
1
5/ Markdown is a human readable markup language, it has been adopted by some note taking and #Zettlekasten tools. But it only covers a subset of the semantics, has been extended inconsistently, and has lots of other deficiencies as outlined by @kmelve
t.ly/iKTd
1
2
8/ Org covers all the listed aspects of #TfT interop, is at least as readable as markdown and has strong developer tool support. Some products already use org-mode outside emacs (@logseq, @plainorg, @aBrainTool) and there are a number of parsers available. t.ly/eOo3
9/ Rather than stay closed, reinvent the wheel or continually extend a partial solution, companies in the #TfT space should look at adopting Org for interoperability between tools and, ideally, as a user-accessible data store. More details here 👇
braintool.org/2022/04/29/Too…
